Andreessen Horowitz said in a Nov. 11, 2026, post and accompanying article that an AI-agent economy is taking shape as software assistants become important internet users and intermediaries between consumers and businesses. The venture firm said the shift could require companies to adopt a business-to-agent, or B2A, model in which agents independently search for products, select services and make purchases on users’ behalf.

“The primary users of the Internet are no longer human. They’re AI agents,” the authors wrote.

According to a16z’s estimates, agents already account for most web traffic, write more than half of new articles published online and generate most of the code that developers send to production environments. The firm described this as an early platform shift in which adoption can be measured but most of the potential business models are not yet visible.

AI agents reshape customer acquisition

Artificial intelligence has already expanded what users can do, a16z said, but changes in product distribution are only beginning. Websites and mobile applications have traditionally helped companies find customers; agents could increasingly evaluate offers and complete tasks without continuous human involvement.

Shopify said traffic from AI sources to stores on its platform rose eightfold in the first quarter of 2026 from a year earlier. The number of orders placed after an AI search increased almost 13-fold.

Human Security reported that traffic from agentic browsers increased eightfold between August and October 2026. Muse generated 60% of that activity, the largest share, while OpenAI products collectively accounted for 30%.

GitHub said that, as of September 2026, more than 40% of pull requests began with a draft created by an AI agent, up from 17% six months earlier.

a16z said traffic from AI services may be more valuable to businesses than conventional referral traffic because users directed by language models show greater engagement and are more likely to make purchases. The firm estimated that, even before fully autonomous agents became widespread, about 11% of consumer purchase journeys in the United States involved an AI interaction before the user visited a merchant’s website.

Potential beneficiaries of the B2A model

In its analysis, a16z identified four areas that could benefit from a shift toward business-to-agent commerce.

The first includes developers of AI models, autonomous assistants and products with agentic capabilities. The firm also identified a separate opportunity around ChatGPT, which it said has 1.2 billion weekly active users.

The second area covers new agent-based products, including autonomous small businesses, marketplaces that use AI intermediaries and automated optimization services. The third includes companies adapting their products for agents. While search engine optimization and generative engine optimization help businesses appear in AI-generated answers, a16z said the agent economy would extend that competition to agents’ evaluation of offers and execution of purchases.

The fourth area is infrastructure, including services for payments, identity verification, fraud prevention and communications.

a16z pointed to software development as an early example of the changes. Supabase reported that AI tools create more than 60% of new databases, while Stripe said agents account for nearly 40% of traffic to its documentation and 70% of command-line interface requests for API resources.

The firm said companies that previously blocked bots because of fraud and chargeback risks may need to reconsider that approach as AI agents become more important sources of traffic and revenue.

Mastercard has opened early access to a payment system for AI agents. U.S. banks including Morgan Stanley, JPMorgan, Bank of America and Wells Fargo are also tightening oversight of third-party AI agents that gain access to customers’ accounts.

Source: Incrypted